DeterminantalPointProcesses.jl

!Disclaimer! This package is based on the work of alshedivat/DeterminantalPointProcesses.jl and aims at keeping this package alive.

An efficient implementation of Determinantal Point Processes (DPP) in Julia.

• Exact sampling [1] from DPP and k-DPP (can be executed in parallel).
• MCMC sampling [2] from DPP and k-DPP (parallelization will be added).
• pdf and logpdf evaluation functions [1] for DPP and k-DPP.

• Exact sampling using dual representation [1].
• Better integration with MCMC frameworks in Julia (such as Lora.jl or AbstractMCMC.jl).
• Fitting DPP and k-DPP models to data [3, 4].
• Reduced rank DPP and k-DPP.
• Kronecker Determinantal Point Processes [5].

Any help on these topics would be highly appreciated

Contributions are sought (especially if you are an author of a related paper). Bug reports are welcome.

[1] Kulesza, A., and B. Taskar. Determinantal point processes for machine learning. arXiv:1207.6083, 2012.

[2] Kang, B. Fast determinantal point process sampling with application to clustering. NIPS, 2013.

[3] Gillenwater, J., A. Kulesza, E. Fox, and B. Taskar. Expectation-Maximization for learning Determinantal Point Processes. NIPS, 2014.

[4] Mariet, Z., and S. Sra. Fixed-point algorithms for learning determinantal point processes. NIPS, 2015.

[5] Mariet, Z., and S. Sra. Kronecker Determinantal Point Processes. arXiv:1605.08374, 2016.